JP2002077723A - Moving image processor and moving image processing method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce arithmetic quantity necessary for the correction of a moving image. SOLUTION: An encoding unit 100 for encoding a moving image is provided with a correcting part 121, a scene judging part 122, a correction data generating part 123, and a scene change detecting part 124. The scene change detecting part 124 detects the scene change of the moving image based on a different picture at the time of encoding, and inputs the detected result to a scene judging part 122. The scene judging part 122 judges the scene of the picture after the scene change by using a predicted picture from the movement compensating part 116 in detail, and a correction data generating part 123 decides a correcting method based on the result of the scene judgment. The correcting part 121 performs correction by the decided correcting method until the next scene change is detected. Thus, the correcting method is updated only when the scene change is generated so that the correction can be properly performed, and that the arithmetic quantity necessary for the correction can be reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for correcting various image characteristics such as gradation, hue, and saturation of a moving image acquired as digital data.

[0002]

2. Description of the Related Art Conventionally, there has been proposed a technique for correcting the gradation and the like of a moving image in real time. For example, Japanese Patent Application Laid-Open No. Hei 5-212620 discloses a moving image processing apparatus that performs simple tone correction in real time. However, the correction of a moving image by the apparatus described in the above-mentioned document is limited, and appropriate correction according to the characteristics of the image of each frame is not performed.

On the other hand, as a correction for a still image, for example, Japanese Unexamined Patent Application Publication No. 2000-57335 discloses a method of obtaining characteristic amounts such as gradation, hue, and saturation of an image, and performing advanced correction based on these characteristic amounts. A technique for performing the method is described.

[0004]

By the way, when an image of each frame of a moving image is treated as a still image and an advanced correction for the still image is applied to the correction of the moving image,
The amount of calculation required for correction increases. To perform processing on a moving image in real time, for example, 30 fps (1
For a moving image of 30 frames per second), the time required for processing one frame needs to be 33 ms or less. Therefore, it is necessary to develop an expensive device to apply the correction method for a still image to real-time correction of a moving image.

For this reason, conventionally, when performing advanced correction on a moving image, the moving image has been once stored on a recording medium such as a so-called hard disk, and then processed in a non-real-time manner. Even in this case, since the amount of calculation required for correcting the entire moving image is enormous, the processing requires a lot of time.

SUMMARY OF THE INVENTION The present invention has been made in consideration of the above problems, and has as its object to reduce the amount of calculation required for moving image correction and to quickly perform moving image correction.

[0007]

According to a first aspect of the present invention, there is provided a moving image processing apparatus, comprising: means for acquiring scene change information indicating a scene change in a moving image; and acquiring the scene change information. Means for determining a method of correcting the moving image until the next scene change information is obtained.

According to a second aspect of the present invention, there is provided the moving image processing apparatus according to the first aspect, wherein the moving image is corrected according to the correction method until the next scene change information is obtained. Is further provided.

According to a third aspect of the present invention, there is provided the moving image processing apparatus according to the first or second aspect, further comprising means for storing a plurality of representative correction methods in advance, and determining the correction method. The means selects one of the plurality of correction methods based on the image after the scene change information is obtained.

According to a fourth aspect of the present invention, in the moving image processing apparatus according to any one of the first to third aspects, the means for acquiring the scene change information includes an image of a frame preceding a current frame. The scene change information is generated based on a difference image between the predicted image of the current frame and the image of the current frame derived from the above.

According to a fifth aspect of the present invention, in the moving picture processing apparatus according to the fourth aspect, the means for determining the correction method determines the correction method based on the predicted image.

According to a sixth aspect of the present invention, there is provided the moving image processing apparatus, wherein the means for acquiring scene change information indicating a scene change in the moving image, and the moving image until the next scene change information is acquired. And a means for correcting the moving image according to the correction method until the next scene change information is obtained.

According to a seventh aspect of the present invention, in the moving image processing apparatus according to the second or sixth aspect, the correction of the moving image is executed in real time.

The invention according to claim 8 is a moving image processing method, wherein a step of acquiring scene change information indicating a scene change in a moving image, and a step of acquiring the moving image until the next scene change information is acquired. And a step of correcting the moving image according to the correction method until the next scene change information is obtained.

According to a ninth aspect of the present invention, there is provided a computer-readable recording medium having recorded thereon a program for causing a computer to execute a moving image correction, wherein the execution of the program by the computer causes the computer to execute the moving image correction. Acquiring scene change information indicating a scene change in, and acquiring the moving image correction method until the next scene change information is acquired, and acquiring the next scene change information until the next scene change information is acquired. Performing the correction of the moving image according to the correction method.

[0016]

DESCRIPTION OF THE PREFERRED EMBODIMENTS <1. First Embodiment> FIG.
1 is a diagram illustrating a configuration of an image processing system 1 that acquires, corrects, and reproduces a moving image. An image processing system 1 includes a video camera 1 that acquires a moving image as digital data.
0, a reproducing apparatus 20 that receives a moving image acquired by the video camera 10 via a recording medium 91 such as a magnetic tape and reproduces the moving image, and a display 30 that displays the reproduced moving image.

FIG. 2 is a block diagram showing a main configuration relating to processing of a moving image in the image processing system 1. As shown in FIG. The image processing system 1 includes an encoding unit 100 and a decoding unit 200. A moving image acquired by the video camera 10 is input to the encoding unit 100 as input image data 8.
Entered as 1. The input image data 81 is encoded (that is, compressed) in the encoding unit 100, and encoded data 82 is output. Encoded data 82
Are input to the decoding unit 200 at the time of reproduction, decoded (that is, decompressed), and output as output image data 83.

The encoding unit 100 and the decoding unit 200 may be provided in either the video camera 10 or the playback device 20 as described later, but in the following description, the video camera 10 is
In the following description, the playback device 20 includes the decoding unit 200.

FIG. 3 is a block diagram showing a configuration of an encoding unit 100 that performs encoding while correcting a moving image in real time. Hereinafter, each configuration of the encoding unit 100 will be described.

The division unit 101 divides an image of each frame of a moving image input as input image data 81 into a plurality of partial images (hereinafter, referred to as “blocks”). For example, a raster scan image is converted into a block scan image in units of 8 × 8 pixel blocks.

The blocks generated by the division unit 101 are sequentially input to the correction unit 121 and corrected, and the subtraction unit 102
Is input to

The subtraction unit 102 generates a difference image (hereinafter, referred to as a “difference block”) between a block after correction from the correction unit 121 and a block of a prediction image (hereinafter, referred to as a “prediction block”) from the motion compensation unit 116 described later. Block ").

The difference block output from the subtraction unit 102 is input to the DCT unit 103. DCT section 103
Is the DCT (Discrete Cosine Transform) for the difference block
To convert the signal in the time domain into DCT coefficients in the frequency domain.

The quantizing section 104 quantizes the DCT coefficients from the DCT section 103, and the encoding section 105 performs variable length encoding on the quantized DCT coefficients and sequentially outputs them as encoded data 82.

The DCT coefficient from the quantization unit 104 is also input to the inverse quantization unit 111, and the inverse quantization unit 111
The DCT coefficient is restored. The inverse DCT unit 112
Generate a difference block from the coefficients.

The restored difference block and the prediction block from the motion compensator 116 are input to the adder 113, and the adder 113 adds these blocks. Thereby, a block of the image (that is, decoded data) on which the correction by the correction unit 121 is reflected is generated. Thereafter, the blocks of the generated image are stored in the frame memory 114.

The frame memory 114 functions as a delay unit for one frame, and sequentially outputs the corrected previous frame image blocks while sequentially storing the corrected current frame image blocks.

The motion vector detecting unit 115 includes the dividing unit 1
The block of the image of the current frame from 01 and the block of the image after the correction of the previous frame from the frame memory 114 are input. The motion vector detection unit 115 detects a motion vector 84 indicating the motion of the subject from these blocks. Although not shown in FIG. 2, the motion vector 8
4 (data) together with the encoded data 82
1 to the decoding unit 200.

The motion compensating unit 116 predicts the current frame image block using the motion vector 84 from the motion vector detecting unit 115 and the image block after correction of the previous frame from the frame memory 114. Accordingly, the motion compensation unit 116 generates a prediction block. Then, the prediction block includes the subtraction unit 102 and the addition unit 1
13 and the scene determination unit 122.

The configuration described above is almost the same as the configuration in normal moving picture compression. Next, the correction unit 12 which is a configuration related to the moving image correction in the encoding unit 100
1. The scene determination unit 122, the correction data generation unit 123, and the scene change detection unit 124 will be described.

The correction unit 121 corrects the block input from the division unit 101 according to a predetermined correction method. The content of correction may be any, but in the present embodiment, the description will be made assuming that correction of contrast and lightness (that is, gradation of pixels) is performed.

The scene determining section 122 includes a motion compensating section 116
Based on the predicted image of the current frame (a group of predicted blocks for one frame), scene information indicating the characteristics of the captured image is generated. Further, a correction parameter serving as a reference for correction is output to the correction data generation unit 123 based on the scene information.

The correction data generation unit 123 determines a correction table indicating the characteristics of the correction in accordance with the correction parameters, and outputs the correction table to the correction unit 121. And
The correction unit 121 performs pixel value conversion of the input block while referring to the correction table.

The scene change detecting section 124 includes the subtracting section 1
02, a difference block is input, and a scene change in a moving image is detected based on a difference image for one frame (that is, a group of difference blocks for one frame). When a scene change is detected, scene change information indicating the scene change is input to the scene determination unit 122.

The scene change information is transmitted to the scene determination unit 122
As described above, the scene determination unit 122 performs a scene determination on the predicted image, the correction parameters are input to the correction data generation unit 123, and the correction table is transmitted from the correction data generation unit 123 to the correction unit 121. Is entered.

That is, the encoding unit 100 determines a correction method each time a scene change is performed in a moving image, and corrects the moving image by this correction method until the next scene change. As a result, the amount of calculation required for the correction can be reduced as compared with the case where the unique correction is performed on the image of each frame. The details of the processing related to the correction will be described later.

As described above, the encoding unit 1
00 is similar to a moving picture coding method such as MPEG, in which a motion vector for each block is detected, a difference block from a motion-compensated adjacent frame is obtained, and then variable-length coding such as Huffman coding is performed. It has become. Then, the encoding unit 100 detects a scene change based on the difference image for one frame, and determines a correction method.

Generally, adjacent frames in a moving image have a high correlation unless a scene change is made, and appropriate correction can be realized even if correction is performed using the same correction method as long as no scene change is made. Is done. Therefore, the encoding unit 100 analyzes the image of the frame after the scene change in detail, derives an appropriate correction method, and uses this correction method until the next scene change.
Advanced correction can be performed with a small amount of calculation.

FIG. 4 is a block diagram showing the configuration of a decoding unit 200 for decoding the encoded data 82 and generating the output image data 83. The decoding unit 200
It has the same configuration as a normal decoding device.

The decoding section 201 performs variable-length decoding on the input coded data 82 to obtain quantized DCT coefficients. The inverse quantization unit 202 obtains an original DCT coefficient from the quantized DCT coefficient. And the inverse DCT section 203
As a result, a difference block is obtained from the DCT coefficients.

The addition unit 204 receives the difference block and the prediction block of the current frame from the motion compensation unit 207, and adds these blocks to generate an image block of the current frame.

The generated blocks are sequentially input to the synthesizing unit 205 and synthesized, and the block scan image in block units is converted into a raster scan corrected image. Then, the generated corrected image is output as output image data 83.

On the other hand, the block generated by the adder 204 is also stored in the frame memory 206, and is used by the motion compensator 207 to generate a prediction block when generating the image of the next frame. As described above, the motion vector 84 is input to the decoding unit 200 together with the encoded data 82, and is used when the motion compensation unit 207 performs motion compensation.

Next, the correction unit 121, the scene determination unit 122, and the correction data generation unit 12 in the encoding unit 100
3 and details of the operation of the scene change detection unit 124 will be described. FIG. 5 is a flowchart showing the flow of processing related to correction in the encoding unit 100.

First, the scene change detection section 124 detects a scene change in a moving image (step S11). The scene change can be regarded as a large change in the image in the moving image, and the scene change detection unit 124 calculates the sum of the pixel values in the difference image for one frame. Specifically, the sum of the pixel values is calculated for the sequentially input difference blocks, and the sum is sequentially added to obtain the sum of the pixel values in the difference image for one frame.

The sum of the pixel values in the difference image is an index value indicating the degree of difference between the image of the previous frame and the image of the current frame. When the sum of the pixel values of the difference image exceeds a predetermined threshold value, Is regarded as having undergone a scene change, and the scene change information is sent to the scene determination unit 122.

If a scene change is detected, the scene determination unit 122 obtains a histogram (brightness histogram) of pixel values of the first predicted image (a group of predicted blocks for one frame) after the scene change is detected (step S).
12). Note that the processing on the pixel value is accurately performed on a value derived from the pixel value (gradation in the present embodiment), but in the following description, the processing will be simply described on the pixel value.

FIG. 6 is a view showing an example of the brightness histogram 7. The brightness histogram 7 is divided into a plurality of ranges as indicated by reference numerals 71 to 76, and a detailed scene determination is performed based on a combination of a sum of variances of pixel values in a plurality of regions, a variance, and the like (step S13). .

The scene determination is a process of determining the state of the image. Specifically, the state of the image is a normal state (normal), a state where the contrast is too strong (high contrast), and a state where the contrast is low (low contrast). contrast),
This is a process for determining whether or not a backlight state, a too bright state (over), or a too dark state (under).

When the scene judgment is completed, the scene judgment unit 1
At 22, the parameters necessary for the correction are acquired according to the result of the scene determination. FIG. 7 is a block diagram showing the transfer of various types of information to the scene determination unit 122 and the correction data generation unit 123. In the scene determination unit 122, the parameter table 852 is stored in a predetermined memory in advance.
By comparing the result of the scene determination derived from the predicted image data 851 that is the predicted image with the parameter table 852, a correction parameter required for correction is obtained.

As exemplified in Table 1, the parameter table 852 is a table showing the correspondence between the results of various scene determinations and the parameters required for correction. Then, the contrast correction level and the brightness correction level are input as correction parameters from the scene determination unit 122 to the correction data generation unit 123 (step S1).
4).

[0052]

[Table 1]

The parameter table 852 is set in advance by the operator via the display unit 152 for displaying information to the operator and the operation unit 153 for receiving input from the operator. FIG. 8 is a diagram showing a display screen when the parameter table 852 is set. That is, for example, the screen illustrated in FIG. 8 is displayed on the display unit 152 that is the display of the video camera 10 illustrated in FIG. 1, and the parameters to be set are highlighted in FIG.
By operating the operation unit 153 such as operation buttons according to the displayed contents, the contents of the parameter table 852 are adjusted via the parameter setting unit 151. This allows the user to set a preferred correction method for each scene characteristic, thereby improving the quality of moving image correction.

As shown in FIG. 7, the correction data generator 123 that has obtained the correction parameters selects a correction table that matches the correction parameters from a plurality of representative correction tables 853 stored in the memory 154 in advance (see FIG. 7). Step S15). By preparing a plurality of correction tables, the amount of calculation required for determining the correction table can be reduced.

FIG. 9 is a diagram illustrating characteristics of the correction table. The horizontal axis corresponds to the pixel value before correction, and the vertical axis corresponds to the pixel value after correction. In FIG. 9, a curve 853 a that is approximately convex upward indicates a correction table that is selected when the image is brightened by correction (that is, when the determination result is “under”), and a straight line 853 b is substantially a line. 9 shows a correction table selected when no correction is performed (that is, when the determination result is “normal”). A curve 853c that is approximately convex downward indicates a correction table selected when the image is darkened by correction (that is, when the determination result is “over”). Although illustration is omitted,
When enhancing the contrast, a curve with a large maximum slope at the center is selected, and when weakening the contrast, a curve with a gentle slope at the center is selected.

The selected correction table 854 is stored in the correction unit 1
Then, the correction unit 121 performs pixel value conversion of the image input according to the correction table for each block (step S16).

By the above processing, the correction method until the next scene change is detected by the scene change detection unit 124 is determined by the correction data generation unit 123.
The correction unit 121 operates until the next scene change is detected.
The moving image is corrected according to the correction table determined by the correction data generation unit 123, that is, by a certain correction method. As a result, appropriate correction can be performed without updating the correction method for each frame image, and appropriate moving image correction can be realized with a small amount of calculation.

Further, by reducing the amount of calculation, it is also possible to perform correction in real time when encoding a moving image.

The encoding unit 100 shown in FIG. 3 uses a differential image at the time of encoding for detecting a scene change. Therefore, a special configuration for generating a differential image for detecting a scene change is provided. There is no need to provide it separately. That is, scene change detection is realized without adding a new frame memory in the encoding unit 100. Furthermore, since a predicted image is used instead of an image to be corrected when determining a scene, a frame memory for separately storing an image to be corrected is not required. As a result, the price of the encoding unit 100 can be reduced.

In the above processing, the image (block) input to the correction unit 121 after the detection of a scene change is the image of the second frame. Applied from the first frame. When displaying a moving image, the image of each frame is displayed only for a short period of time. Therefore, even if appropriate correction is not performed only for the first frame, appropriate correction is performed for the entire moving image. Will be done.

If it is necessary to perform appropriate correction from the first frame after the scene change, a frame memory is separately provided upstream of the correction unit 121. Then, a difference image between the image of the previous frame and the image of the current frame is obtained before the correction, and a scene change is detected based on the difference image, whereby appropriate correction is realized from the first frame. In this case, it is also possible to input the image of the first frame after the scene change, instead of the predicted image, to the scene determination unit 122 to perform the scene determination.

<2. Second Embodiment> In the first embodiment, the moving image is corrected by the encoding unit 100. However, the moving image is corrected by the decoding unit 20.
0 can also be performed.

FIGS. 10 and 11 are block diagrams showing the configurations of an encoding unit 100 and a decoding unit 200 according to the second embodiment, respectively. The encoding unit 100 generates correction parameters corresponding to a scene change. And the decoding unit 200 performs the correction based on the correction parameter in real time.

The coding unit 100 has a configuration in which the correction unit 121 and the correction data generation unit 123 are omitted from the coding unit in the first embodiment, and the correction parameters obtained by the scene determination unit 122 are used. 85 is
The encoded data 82 and the motion vector 84 are passed to the decoding unit 200 via the recording medium 91.

The decoding unit 200 includes the adding unit 204 and the combining unit 20 of the decoding unit according to the first embodiment.
5, a correction unit 211 is provided, and a correction data generation unit 213 is connected to the correction unit 211. Then, the correction unit 21
1 and the correction data generation unit 213 perform the same processing as the corresponding configuration in the first embodiment. That is, the correction parameter 85 from the encoding unit 100 is input to the correction data generator 213, and the correction data generator 213 selects a correction table. The selected correction table is input to the correction unit 211, and the correction unit 211 converts the pixel value of the block generated by decoding with reference to the correction table.

The correction parameter 85 is input in synchronization with the input of the encoded data 82 in accordance with a scene change, and the correction method in the correction unit 211 is changed according to the scene change of the moving image.

As described above, it is also possible for the coding unit 100 to make a decision on the correction for a moving image and to make a correction in the decoding unit 200. Even in this case, the correction method can be changed according to a scene change in the moving image, and the amount of calculation required for correcting the moving image can be reduced.

Further, by reducing the amount of calculation, it is possible to perform correction in real time when decoding a moving image. As a result, the price and size of the device can be reduced.

Further, in the second embodiment, the correction parameter 85 and the encoded data
Since the data is transferred to 00 and the correction is performed at the time of decoding, it is easy to select whether or not to perform the correction as needed at the time of decoding.

Further, in the first embodiment, a new correction method is applied from the image of the second frame, but in the second embodiment, the input of the correction parameter 85 is changed to a scene change. By performing it together, it is easy to apply a new correction method from the first frame after the scene change.

<3. Third Embodiment> Next, an embodiment in which all the processing related to the correction is performed in the decoding unit 200 will be described. FIG. 12 and FIG. 13 are block diagrams showing the configurations of the encoding unit 100 and the decoding unit 200 in the case where all the processing relating to the correction is performed by the decoding unit.

The encoding unit 100 has a configuration in which the scene determination unit 122 and the scene change detection unit 124 are further omitted from the encoding unit in the second embodiment, and performs only encoding of a moving image.

The decoding unit 200 is different from the decoding unit in the second embodiment in that
2 and a scene change detection unit 214 are further added, and corrects a moving image in real time while decoding the encoded data 82. That is, the scene change detection unit 214 detects a scene change based on the difference image (a group of difference blocks for one frame) output from the inverse DCT unit 203, and the detection result is input to the scene determination unit 212. In the scene determination unit 212,
Scene determination is performed based on a predicted image (a predicted block group for one frame) from the motion compensation unit 207, and a correction parameter is obtained by referring to a parameter table based on the determination result.

Thereafter, the correction data generation unit 213 selects a correction table most suitable for correction from the plurality of correction tables using the correction parameters, and performs block correction by the correction unit 211 using the selected correction table. .
As a result, the corrected moving image is output from the synthesizing unit 205 as the output image data 83.

As described above, it is also possible for the decoding unit 200 to perform all of the processing relating to the correction of a moving image. Even in this case, it is possible to change the correction method according to the scene change in the moving image,
The amount of calculation required for correcting a moving image can be reduced.

Further, by reducing the amount of calculation, it is possible to perform correction in real time when decoding a moving image.

The decoding unit 200 shown in FIG.
By using the differential image at the time of decoding for detecting a scene change, the addition of a new correction frame memory in the decoding unit 200 is omitted, and the price of the decoding unit 200 is reduced. .

In the third embodiment, a new correction method is applied from the second frame after the scene change, but this does not cause a problem in the whole moving image.

<4. Fourth Embodiment> The encoding unit 100 and / or the decoding unit 200 described in the first to third embodiments can be realized by software using a computer. FIG.
4 connects the video camera 10 and the computer 40,
FIG. 2 is a diagram showing a system in which a computer 40 executes moving image correction.

As shown in FIG. 15, the computer 40 includes a CPU 401 for performing various arithmetic processing, a ROM 402 for storing a basic program, and an RA for storing various information.
This is a general computer system configuration in which M403 is connected to a bus line. In addition to the bus line,
A fixed disk 404 for storing a large amount of information, a display 405 for displaying images, a keyboard 406a and a mouse 406b for receiving input from an operator, and recording media 9 such as an optical disk, a magnetic disk, and a magneto-optical disk.
A reading device 407 for reading information from 2, and
Communication unit 408 for receiving a signal from video camera 10
Are appropriately connected via an interface (I / F) or the like.

When the moving image processing is executed by the computer 40, the moving image processing program 441 is read from the recording medium 92 via the reading device 407 in advance and stored in the fixed disk 404. . Then, the program 441 is copied to the RAM 403 and C
The moving image processing is realized by the PU 401 executing arithmetic processing according to the program 441 in the RAM 403. At this time, various information and moving images are displayed on the display 405 as necessary. Note that the keyboard 406
a and the mouse 406b are used for setting the parameter table.

When the computer 40 is used as the encoding unit 100 and the decoding unit 200 according to the first embodiment, an image signal from the video camera 10 is input as digital data via the communication unit 408,
The CPU 401 or the like in the computer 40 performs the same processing as the various configurations shown in FIG.
4 stores the corrected encoded data 82 (and the motion vector 84). When reproducing a moving image, the CPU
The moving image is displayed on the display 405 by performing processing similar to the various configurations illustrated in FIG.

If the encoding process cannot be performed in real time due to the performance of the CPU 401 or the like, the moving image data is temporarily stored on the fixed disk 404, and then the encoded data 82 is generated.

Similarly, when the second and third embodiments are implemented by a computer 40, the computer 40 is caused to function as the encoding unit 100 and the decoding unit 200 shown in FIGS.

Note that the computer 40 includes first to third
The operation of only one of the encoding unit 100 and the decoding unit 200 according to the embodiment may be realized. For example, the decoding unit 2 according to the third embodiment
When only 00 is realized by the computer 40, a device that performs only normal encoding processing is used as the video camera 10, and a moving image is corrected when the computer 40 performs decoding.

As described above, the first to third embodiments can be realized using a computer. Even in this case, the amount of calculation can be reduced.
It is possible to quickly perform processing on a moving image.

<5. Modifications> While the preferred embodiments of the present invention have been described above, the present invention is not limited to the above-discussed preferred embodiments, but allows various modifications.

For example, in the above embodiment, the encoding unit 100 and / or the decoding unit 200 perform the processing related to the correction of the moving image. Regardless, a scene change may be detected and the moving image may be corrected.

In the above embodiment, the subtraction unit 102 of the encoding unit 100 and the inverse D
Although a difference image is acquired from the CT unit 203 to detect a scene change, the detection of a scene change may be performed by another method. For example, a scene change may be simply detected from a difference image between the image of the previous frame and the image of the current frame. Further, a scene change may be detected from a histogram of pixel values of a difference image or a motion vector.

The information indicating the timing of the scene change may be prepared separately from the moving image. That is, the scene change does not need to be detected from the moving image, and the information indicating the scene change may be acquired by separately inputting the information.

Further, in the above embodiment, a large change in a moving image is detected as a scene change. In this case, the detected scene change does not always coincide with a physical scene change at the time of shooting. . If a moving image changes greatly even in one scene, it is detected as a scene change. As described above, the scene change in the above description does not need to coincide with the change of the physical scene, and the change of the moving image corresponding to the scene change is detected as the scene change. In addition, the amount of calculation can be reduced.

In the above embodiment, the correction is performed on the image of each frame. However, the correction may be performed on the difference image. In this case, a correction table for the difference image is used. When the correction is performed on the difference image, the correction of the moving image is performed after the subtraction unit 102 of the encoding unit 100 or the inverse DCT unit 2 of the decoding unit 200.
03 can also be performed. As described above, the correction of a moving image based on a scene change can be performed at an arbitrary stage.

Further, in the above-described embodiment, it has been described that the encoded data 82 is input from the video camera 10 to the reproducing device 20 via the magnetic tape, but any method can be used as a data transfer method. May be done. For example,
An IC memory or a recording disk may be used as a recording medium for transfer, or wireless communication or wired communication via a transmission cable or a computer network may be used. Note that various techniques may be similarly used for data transfer between the video camera 10 and the computer 40 in the fourth embodiment.

In the first to third embodiments, the encoding unit 100 is provided in the video camera 10 and the decoding unit 200 is provided in the reproducing device 20. Video camera 1
0, or may be provided in the playback device 20.
That is, the video camera 10 and the playback device 20 in the above description are only specific examples, and various configurations of the encoding unit 100 and the decoding unit 200 may be provided in any manner.

Further, in the above-described embodiment, it has been described that image data is input from the video camera 10 to the reproducing device 20 or the computer 40. However, instead of the video camera 10, another image output device such as a video deck is used. May be used.

In the above-described embodiment, the motion compensation is performed to obtain the prediction block of the current frame, the movement of the main subject in the same scene is prevented from being detected as a scene change, and the frequency of scene change detection is reduced. In this way, the amount of calculation is reduced. However, the configuration related to the motion compensation may be omitted for applications other than those with little motion such as fixed-point observation (for example, monitoring by a monitoring camera). In this case, the image of the previous frame is used as the predicted image. Note that the predicted image is not limited to the image after the motion compensation or the image of the previous frame, but may be any image that is derived from the image before the previous frame and can be used as the predicted image of the current frame. May be used.

Further, the correction in the above embodiment is not limited to the correction of the luminance gradation, but the correction of another image characteristic amount such as saturation, hue, color saturation, or a plurality of image characteristic amounts. It may be.

In the above embodiment, the dividing unit 101
Although the image of each frame is divided into blocks in, processing such as correction, encoding, and decoding may be performed without dividing the image into blocks. Conversely, in the above embodiment, the same correction is performed for all the blocks of one frame, but correction using a different correction table may be performed for each block. In this case, a correction table corresponding to each block is used by the correction unit, and when a scene change is detected, these correction tables are updated.

In the above-described embodiment, the presence or absence of processing in the correction data generation unit or the like is determined according to the presence or absence of detection in the scene change detection unit, and the processing time changes. Therefore, in order to perform processing more quickly, the rate of the data amount input to the correction unit may be changed according to the presence or absence of a scene change. For example, as shown by a broken line in FIG. 3, a scene change detection result is input from the scene change detection unit 124 to the division unit 101,
The data transfer rate may be increased while a scene change is not detected, and may be decreased when a scene change is detected.

Further, in the above embodiment, it has been described that a plurality of representative correction tables are prepared in advance as a plurality of correction methods. However, the correction table may be generated each time a scene change is detected. For example, a correction table may be generated from a cumulative histogram of pixel values in a predicted image or a decoded image. That is, the correction table may be generated by performing normalization of the frequency value, clipping of a value equal to or more than a certain value, addition of a certain value, correction of a black / white edge, and the like for the cumulative histogram.

In the fourth embodiment, the encoding unit 100 and / or the decoding unit 200 are realized by using the computer 40. However, a part of the encoding unit 100 and / or the decoding unit A part of the unit 200 may be realized by a computer. The various configurations in the first to third embodiments do not need to be clearly separated from each other in terms of hardware, and may be realized using a logic circuit or a microcomputer as appropriate. For example, the processing in the correction data generation unit may be realized by a microcomputer, and the other configuration may be realized by a logic circuit. Further, the encoding unit 100 and / or the decoding unit 200 may be constructed by a plurality of computers.

Further, as shown in the first to third embodiments, the configuration relating to the correction can be arbitrarily provided separately in the encoding unit 100 and the decoding unit 200. For example, it is also possible to provide only the scene determination unit 122 in the encoding unit 100 and provide other components related to correction in the decoding unit 200.

[0103]

According to the first to ninth aspects of the present invention, it is possible to reduce the amount of calculation required for correcting a moving image.

According to the third aspect of the present invention, the amount of calculation required to determine the correction method is reduced.

According to the fourth aspect of the present invention, scene change information can be appropriately acquired, and further, a difference image acquired at the time of encoding or decoding of a moving image can be used. .

According to the fifth aspect of the present invention, the correction method is determined not by using the image to be corrected but by using the predicted image used to obtain the difference image. Therefore, the image to be corrected is separately stored. No means is required.

According to the sixth aspect of the present invention, it is easy to correct a moving image in real time.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a configuration of an image processing system.

FIG. 2 is a block diagram showing a main configuration relating to processing of a moving image.

FIG. 3 is a block diagram illustrating a configuration of an encoding unit.

FIG. 4 is a block diagram illustrating a configuration of a decoding unit.

FIG. 5 is a diagram showing a flow of processing relating to correction.

FIG. 6 is a diagram illustrating a histogram of a predicted image.

FIG. 7 is a block diagram illustrating transfer of various types of information to a scene determination unit and a correction data generation unit.

FIG. 8 is a diagram showing a display screen when setting a parameter table.

FIG. 9 is a diagram illustrating characteristics of a correction table.

FIG. 10 is a block diagram illustrating a configuration of an encoding unit.

FIG. 11 is a block diagram illustrating a configuration of a decoding unit.

FIG. 12 is a block diagram illustrating a configuration of an encoding unit.

FIG. 13 is a block diagram illustrating a configuration of a decoding unit.

FIG. 14 is a diagram illustrating a system that executes correction of a moving image by a computer.

FIG. 15 is a block diagram illustrating a configuration of a computer.

[Explanation of symbols]

 Reference Signs List 10 video camera 20 playback device 40 computer 92 recording medium 100 encoding unit 121, 211 correction unit 122, 212 scene determination unit 123, 213 correction data generation unit 124, 214 scene change detection unit 154 memory 200 decoding unit 441 program 851 prediction Image data S11 to S16 Step

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04N 5/20 H04N 5/20 5C066 5/222 5/222 Z 5C082 7/32 9/64 A 9/64 9/68 A 9/68 7/137 ZF term (reference) 5B057 BA19 CA08 CA17 CB08 CB17 CE11 CE16 DB06 5C021 PA58 PA62 PA66 PA77 PA79 PA80 RA01 RB00 RB06 XA13 XA35 5C022 AC01 AC69 CA00 5C023 AA01 AA04 BA08A EA13 5C059 KK01 LA00 MA05 MA05 NN01 NN21 NN43 PP04 PP14 SS14 SS20 TA01 TC14 TD05 TD10 UA38 5C066 AA01 CA08 EA03 EA05 EB01 EC01 EF00 EF16 KD06 KE02 KE09 KG01 KP02 5C082 AA27 BB12 BA35 BA35 BA35 BA35 BA35 BA35 BA35

Claims (9)

[Claims] 1. A moving image processing apparatus, comprising: means for acquiring scene change information indicating a scene change in a moving image; and when the scene change information is acquired, until the next scene change information is acquired. Means for determining the method of correcting the moving image. 2. The moving image processing apparatus according to claim 1, further comprising: means for correcting the moving image according to the correction method until the next scene change information is obtained. Moving image processing device. 3. The moving image processing apparatus according to claim 1, further comprising: means for storing a plurality of representative correction methods in advance, wherein the means for determining the correction method includes acquiring scene change information. A moving image processing apparatus, wherein one of the plurality of correction methods is selected based on a later image. 4. The moving image processing apparatus according to claim 1, wherein the means for acquiring the scene change information predicts a current frame derived from an image of a frame earlier than a current frame. A moving image processing apparatus, wherein the scene change information is generated based on a difference image between an image and an image of a current frame. 5. The moving image processing apparatus according to claim 4, wherein the means for determining the correction method determines the correction method based on the predicted image. 6. A moving image processing apparatus, comprising: means for acquiring scene change information indicating a scene change in a moving image; and means for acquiring a method for correcting the moving image until the next scene change information is acquired. And a means for correcting the moving image according to the correction method until the next scene change information is obtained. 7. The moving image processing apparatus according to claim 2, wherein the correction of the moving image is performed in real time. 8. A moving image processing method, comprising: obtaining scene change information indicating a scene change in a moving image; and obtaining a moving image correction method until the next scene change information is obtained. And a step of correcting the moving image according to the correction method until the next scene change information is obtained. 9. A computer-readable recording medium having recorded thereon a program for causing a computer to execute a moving image correction, wherein the execution of the program by the computer causes the computer to execute a scene change indicating a scene change in a moving image. Obtaining information, and obtaining a correction method of the moving image until the next scene change information is obtained; and until the next scene change information is obtained, obtaining the correction method of the moving image according to the correction method. Performing a correction step.